Comments on: SIGIR ’09 Accepted Papers http://thenoisychannel.com/2009/04/22/sigir-09-accepted-papers/ Mon, 21 May 2012 05:21:00 +0000 hourly 1 http://wordpress.org/?v=3.3.2 By: Daniel Tunkelang http://thenoisychannel.com/2009/04/22/sigir-09-accepted-papers/comment-page-1/#comment-2958 Daniel Tunkelang Thu, 23 Apr 2009 02:02:53 +0000 http://thenoisychannel.com/?p=2000#comment-2958 A while back I talked with Harr Chen and David Karger about their "<a href="http://people.csail.mit.edu/harr/papers/sigir2006.pdf" rel="nofollow">Less is More</a>" work and suggested it might be valuable to return a set of results where exactly k of them are relevant, rather than at least k (they propose a "k-call" measure). I argued based on the paradox of choice: users want a small set of plausible options, and might actually want to see a few other options that are clearly less desirable. They were intrigued, but we never followed up on the idea. A while back I talked with Harr Chen and David Karger about their “Less is More” work and suggested it might be valuable to return a set of results where exactly k of them are relevant, rather than at least k (they propose a “k-call” measure). I argued based on the paradox of choice: users want a small set of plausible options, and might actually want to see a few other options that are clearly less desirable. They were intrigued, but we never followed up on the idea.

]]> By: jeremy http://thenoisychannel.com/2009/04/22/sigir-09-accepted-papers/comment-page-1/#comment-2954 jeremy Thu, 23 Apr 2009 00:31:27 +0000 http://thenoisychannel.com/?p=2000#comment-2954 <i>No papers from current UMass researchers (though a number from alumni, and decent representation in the posters)</i> There was a huge turnover in the UMass program in the last few years, with most of the senior grad students leaving from 2006-2008. In the current batch of students, most if not all are < 2 years into the program. I think that explains why there are no UMass papers, more than any budget cuts. The current crop of students hasn't been around long enough to really build up their research yet. <i>When More Is Less: The Paradox of Choice in Search Engine Use</i> Hey, that does sound interesting. I wonder what tack the authors will take. Usually when I hear the paradox of choice quoted in reference to search engines, it is to make the argument that faceted search, or exploratory search, or any number of helpful assistance gives the users too much choice, and therefore are to be avoided. My perspective is the opposite. When all the search engine offers you is a small blank box, you basically have trillions of choices at your disposal, trillions of decisions to make: What query are you going to enter? How long is it going to be? What query modifiers will you use? The choices are, IMO, overwhelming. The command line offers way too much choice. On the other hand, by giving the user 10 or 20 or even 50 facets or query modifier suggestions, etc. you are narrowing those trillions of choices down to 1o or 20 or 50. It should be much, much easier, then for the user to use the search engine. That's just my opinion, though. I'm looking forward to what the authors of the paper have to say, and what experiments they've done. BTW, I make a similar argument here: <a href="http://irgupf.com/2009/03/04/ranked-lists-and-the-paradox-of-choice/" rel="nofollow">http://irgupf.com/2009/03/04/ranked-lists-and-the-paradox-of-choice/</a> No papers from current UMass researchers (though a number from alumni, and decent representation in the posters)

There was a huge turnover in the UMass program in the last few years, with most of the senior grad students leaving from 2006-2008. In the current batch of students, most if not all are < 2 years into the program. I think that explains why there are no UMass papers, more than any budget cuts. The current crop of students hasn’t been around long enough to really build up their research yet.

When More Is Less: The Paradox of Choice in Search Engine Use

Hey, that does sound interesting. I wonder what tack the authors will take. Usually when I hear the paradox of choice quoted in reference to search engines, it is to make the argument that faceted search, or exploratory search, or any number of helpful assistance gives the users too much choice, and therefore are to be avoided.

My perspective is the opposite. When all the search engine offers you is a small blank box, you basically have trillions of choices at your disposal, trillions of decisions to make: What query are you going to enter? How long is it going to be? What query modifiers will you use? The choices are, IMO, overwhelming. The command line offers way too much choice.

On the other hand, by giving the user 10 or 20 or even 50 facets or query modifier suggestions, etc. you are narrowing those trillions of choices down to 1o or 20 or 50. It should be much, much easier, then for the user to use the search engine.

That’s just my opinion, though. I’m looking forward to what the authors of the paper have to say, and what experiments they’ve done.

BTW, I make a similar argument here: http://irgupf.com/2009/03/04/ranked-lists-and-the-paradox-of-choice/

]]>